The characteristics of the O
2
Herzberg II and Chamberlain bands observed
with VIRTIS/Venus Express
A. Migliorini
a,⇑, G. Piccioni
a, J.C. Gérard
b, L. Soret
b, T.G. Slanger
c, R. Politi
a, M. Snels
d, P. Drossart
e,
F. Nuccilli
aa
IAPS-INAF, via del Fosso del Cavaliere, 100, 00133 Rome, Italy
bLPAP, ULg, Allée du 6 Août, 17 – Sart Tilman, B-4000 Liège, Belgium c
MPL, SRI International, Menlo Park, CA, United States
d
ISAC/CNR, via del Fosso del Cavaliere, 100, 00133 Rome, Italy
e
LESIA, 5, place Jules Janssen, 92195 Meudon, Paris, France
a r t i c l e
i n f o
Article history: Available online xxxx Keyword: Venus Venus Atmosphere Spectroscopya b s t r a c t
The oxygen Venus nightglow emissions in the visible spectral range have been known since the early observations from the Venera spacecraft. Recent observations with the VIRTIS instrument on board Venus Express allowed us to re-examine the Herzberg II system of O2and to further study its vertical
distribu-tion, in particular the (0–m00withm00= 7–13) bands. The present work describes the vertical profile of the observed bands and relative intensities from limb observation data. The wavelength-integrated intensi-ties of the Herzberg II bands, withm00= 7–11, are inferred from the recorded spectra. The resulting values lie in the range of 84–116 kR at the altitudes of maximum intensity, which are found to lie in the range of 93–98 km.
Three bands of the Chamberlain system, centered at 560 nm, 605 nm, and 657 nm have been identified as well. Their emission peak is located at about 100 km, 4 km higher than the Herzberg II bands.
For the first time, the O2nightglow emissions were investigated simultaneously in the visible and in
the IR spectral range, showing a good agreement between the peak position for the Herzberg II and the O2ða1Dg—X3RgÞ bands. An airglow model, proposed by Gérard et al. (Gérard, J.C., Soret, L., Migliorini, A., Piccioni, G. [2012]. Icarus.) starting from realistic O and CO2vertical distributions derived from
Venus-Express observations, allows reproduction of the observed profiles for the three O2systems.
Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction
The oxygen emissions in the 300–700 nm spectral range, known as the O2ðc1Ru—X
3
R
gÞ Herzberg II bands, were observed for the
first time in the Venera 9 and 10 data (Krasnopolsky et al., 1977). Successive laboratory experiments (Lawrence et al., 1977; Slanger, 1978) made this assignment. Moreover, a second weaker progres-sion was recognized in the Venera 9 data, and assigned to the O
2-(A03D
u–a1Dg) Chamberlain system (Slanger and Black, 1978).
The reported intensities for these two band systems were 2– 3 kR (1R = 106photon cm2s1 (4
p
ster)1) for the entireO2ðc1Ru—X 3R
gÞ 0–
m
00 progression in the Venera measurementsand about 15 times less for the O2(A03Du–a1Dg) 0–
m
00 progression(Krasnopolsky, 1983). Similar values for the Herzberg II band
pro-gression were reported byBougher and Borucki (1994)from the Pioneer Venus Orbiter data.
From ground-based observations, a few rotational lines belong-ing to the (0–10) Herzberg II band were identified in observations with the Keck I and Apache Point Observatory telescopes by Slan-ger et al. (2001, 2006), during an observation campaign to detect the O1S–1D green line of atomic oxygen.
More recently, a Herzberg II series of bands was observed using the visible channel of VIRTIS (Visible and InfraRed Thermal Imag-ing Spectrometer), the imagImag-ing spectrometer on board the Euro-pean mission to Venus, Venus Express. By averaging thousands of spectra acquired using the instrument in limb-mode observation, the 0–
m
00 progression was clearly detected on the night side ofthe planet (García-Muñoz et al., 2009). A total limb intensity of 128.4 kR was reported by the authors for the summed intensity of the
m
00= 6–11 bands. Due to the weakness of these bands, thevertical profile was difficult to derive in the available data. How-ever it was pointed out that the emission intensity was stronger at about 95 ± 1 km height, as shown in Figs. 4 and 5 in García-Muñoz et al. (2009).
0019-1035/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.icarus.2012.11.017 ⇑ Corresponding author. Fax: +39 0645488188.
E-mail addresses: Alessandra.Migliorini@iaps.inaf.it(A. Migliorini), Giuseppe. Piccioni@iaps.inaf.it (G. Piccioni), JC.Gerard@ulg.ac.be (J.C. Gérard), lauriane. soret@ulg.ac.be (L. Soret), tom.slanger@sri.com (T.G. Slanger), Romolo.Politi@ iaps.inaf.it (R. Politi), m.snels@isac.cnr.it (M. Snels), pierre.drossart@obspm.fr
(P. Drossart),Fabrizio.Nuccilli@iaps.inaf.it(F. Nuccilli).
Contents lists available atSciVerse ScienceDirect
Icarus
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / i c a r u s
The Herzberg II band system is not detected in the high resolu-tion nightglow spectra of the Earth (Cosby et al., 2006), while emis-sion lines of the c ? b band system, which originates from the O2ðc1RuÞ state, are detected (Slanger et al., 2003). By combining
laboratory studies and airglow observations of the terrestrial plan-ets, it was deduced that there exists a strict dependence between the ðc1R u—X 3 RgÞHerzberg II/ðA 3 Rþu—X 3
RgÞHerzberg I and the
[M]/[O] ratios (Slanger and Copeland, 2003). In the case of Venus and Mars M is CO2, while for the Earth it is N2. According to the
concentrations of CO2, N2and O in the atmospheres of the three
terrestrial planets at the nightglow altitudes, the [M]/[O] ratio ex-plains the presence of an intense Herzberg II emission on Venus and its absence on the Earth (Greer and Murtagh, 1985; Slanger and Copeland, 2003).
Although the ðA3Rþ u—X
3R
gÞ Herzberg I and ðA 03D
u X3RgÞ
Herzberg III systems are not ruled out in the Venus’ atmosphere, the expected intensities are very low. The Herzberg I system is ex-pected to be about 20 times less intense than the Herzberg II in the Venus’ atmosphere, according toKrasnopolsky (1983). Considering the VIRTIS spectral sampling, it is hard to separate them from the Herzberg II bands. Hence it is not possible to unambiguously iden-tify any feature belonging to the Herzberg I and III systems from VIRTIS data. In addition, at wavelengths shorter than 400 nm the stray light contribution is strong and the instrument sensitivity is not sufficient for weak band detection. As a consequence, the spec-tral region 300–400 nm must be treated with caution.
An extension of the previous work by García-Muñoz et al. (2009)concerning the O2nightglow emissions in the Venus
atmo-sphere is provided in the present paper by using VIRTIS/Venus Ex-press data acquired during a recent observing campaign at limb. Details of selected data, and data processing are reported in Sec-tion2. Using the new observing mode, a better coverage of the alti-tude region around 95 km, where the Herzberg II band emissions are known to occur, is obtained. It has also allowed derivation of the vertical profile for the detected bands. In addition, for limb data acquired in March 2007 and April 2008, it was possible to simulta-neously investigate the nightglow O2emissions in the visible and
the IR Atmospheric band at 1.27
l
m. Results are discussed in Section3.2. Observations with VIRTIS-M on Venus Express
The VIRTIS (Visible and InfraRed Thermal Imaging Spectrome-ter) instrument on board Venus Express is an imaging spectrome-ter composed of two channels. One, called M, covers the range from 0.3 to 5.1
l
m using 864 bins, called bands, and has imaging capabilities, while the second, called H, has a higher spectral reso-lution but no imaging capabilities. In this study, we describe re-sults obtained with data acquired with the M-channel of VIRTIS. The M-channel can cover the visible spectral range from 0.3 to 1.1l
m by steps of about 2 nm, while the IR part is sampled with 10 nm steps. The VIRTIS-M spectrometer can operate both in na-dir and limb modes. Because of the orbital configuration, the nana-dir mode observation is typically used in the ascending branch of the orbit, while the limb observations are concentrated when the spacecraft is near pericenter. As a consequence, the southern hemi-sphere of Venus is better mapped with nadir observations, while the limb observations preferentially sound the northern hemi-sphere. A detailed description of the instrument is reported in Drossart et al. (2007) and Piccioni et al. (2007).For faint emissions such as those discussed in the present work, limb mode observations are the most suitable, as the signal is amplified by a factor of 50 compared to nadir observations be-cause of the long path in the line of sight through the atmosphere. A campaign of limb observations has been recently conducted,
using the so-called ‘‘limb-tracking’’ technique. This technique takes advantage of the VIRTIS internal mirror capability to span the ver-tical direction. The z-axis of the instrument points at a fixed alti-tude, usually at 100 km or 120 km, depending on the distance of the spacecraft from the Venus north pole, and the VIRTIS image is built up by a sequence of scans at different altitudes, using the motion of the internal mirror of the spectrometer. The major inno-vation is that the limb altitude is kept fixed while the spacecraft moves along its orbit. This makes it possible to obtain a better sam-pling of a restricted altitude region, although the coverage of each single image is limited in latitude and local time. Data with an exposure time of 18 s have been used, in order to have a good sig-nal to noise ratio (SNR). Details of the data are reported inTable 1. The available data cover a restricted area around the anti-solar point as a consequence of the limb-tracking observing mode, as ex-plained above. Owing to the limited coverage, an investigation of the O2emissions behavior with latitude cannot be carried out at
this stage in the visible spectral range but the observations are pro-gressing in the course of the mission.
Data have been calibrated applying a pipeline consolidated dur-ing the 5 years of the mission and outlined below. Data, originally in digital number (DN), are first of all initialized in order to identify the best parameters for the subsequent processes. Dark frames, ac-quired each 20 lines in a scientific image, are identified, analyzed in order to obtain the thermal evolution law of the observations, and removed. Then corrections for thermal evolution, bad frames, dead and saturated pixels, all known as the pre-processing stage, are ap-plied. In the calibration stage, digital numbers are converted to radiance using the VIRTIS Instrument Transfer Function (ITF), and finally a post-processing stage is applied, in order to remove spikes and attenuate the detector non-uniformity response due to differ-ences of responsivity between adjacent rows or columns. This stage includes also the spectral registration, which allows assign-ing a wavelength value to each band, accordassign-ing to a linear relation between wavelength and detector temperature derived from the internal calibration of the instrument, acquired during different mission phases (Cardesín Moinelo et al., 2010).
About 8700 spectra in the altitude range 90–120 km were se-lected in total. The spectrum, obtained by averaging the sese-lected spectra, is shown inFig. 1.
The Herzberg II is the prominent system observed on the night side upper mesosphere of Venus in the visible range. Eight bands of the (0–
m
00) progression clearly stand out above the noise; the (0–6)band at 422 nm, (0–7) at 450 nm, (0–8) at 480 nm, (0–9) at 514 nm, (0–10) at 552 nm, (0–11) at 594 nm, (0–12) at 641 nm, and (0–13) at 702 nm. The continuum is estimated to be less than 10% of the band intensities. In addition, three bands, centered at 558 nm, 604 nm, and 657 nm are also observed close to the Herz-berg II bands. They have been identified as the (0–6), (0–7), and (0– 8) transitions of the Chamberlain system.
In the following section, we discuss the vertical profile and intensity for the most intense bands of the Herzberg II and Cham-berlain systems.
3. Results
An estimate of the tangent altitude of the Herzberg II bands was reported byGarcía-Muñoz et al. (2009), although a direct measure of individual limb profiles was quite difficult because of the low SNR. Owing to the refined limb observing technique with VIRTIS, described above, we were able to derive the tangent profile of the emissions.
In the present study, the (0–7), (0–8), (0–9), (0–10), and (0–11) Herzberg II bands are considered. The other bands, though unam-biguously detected on the averaged spectrum, have a low SNR
which does not allow further investigation. For each band, the radi-ance was first converted into Rayleigh units (Baker, 1974). The radiance was then integrated considering the wavelength ranges 446–459 nm for the (0–7) band, 476–484 nm for (0–8), 511– 518 nm for (0–9), 548–556 nm for (0–10), and 592–596 nm for the (0–11) band, respectively. The continuum was estimated fol-lowing the method described inSoret et al. (2010). The continuum is likely due to infield stray light due to scattering of thermal emis-sion by haze from nearby atmospheric windows. To remove this contribution, a first order polynomial fit is applied above 105 km and below 90 km.Soret et al. (2010)argued that the applied meth-od works well for bright emissions, while the continuum contribu-tion is difficult to estimate in the case of weak emissions. In the visible spectral range where the detected emissions are weak, we consider the integrated emission of the Herzberg II system, instead of using single bands, in order to have a higher SNR and hence a better definition of the emission. The boundaries at 105 and 90 km do not work well with the Chamberlain bands, which are fainter than the Herzberg II bands; hence, ad hoc boundaries were adopted in this case. Spectra were selected in the altitude range 90–120 km, for each acquired VIRTIS image, and divided into 3-km bins. Spectra in each altitude bin were averaged and displayed versus altitude.Fig. 2illustrates examples of the mean intensity distribution for four observed limb profiles. These images have been selected in different periods of the mission, in order to verify also the variability of the Herzberg II bands with time.
During these observations, the spatial resolution varied from 1 km, in the limb-tracking mode, to 2.5 km for the tangential limb observations. In order to avoid spurious features in the case of
weak emissions and have an homogeneous vertical resolution, we use a bin resolution of 3 km for all the analyzed data. This is also the uncertainty associated with the peak altitude derived for the tangent vertical profiles.
In addition, we consider the limb profile of the integrated band system, as we previously verified that each band peaked at the same altitude, as required, within the estimated errors.The limb profiles were then fitted with a Gaussian curve to estimate the Table 1
List of VIRTIS data used in the present work.
Limb session Jtime Latitude Local time Peak (km) Intensity (kR) FWHM (km)
1288-05 () 2455134.6 0.5–13.5S 23–0.2 h 94.9 76 1290-04 ( ) 2455136.6 0.5–13.5S 23.5–0.5 h 94.2 132 1292-04 () 2455138.6 0–13.5S 23.7–0.6 h 96.1 89 10.2 1374-07 () 2455220.8 4–14.5S 20.1–20.8 h 94.8 76 11.5 1375-06 2455221.8 3.5–14.5S 20–21 h 93.0 116 12.7 1377-03 () 2455223.8 3.5–14.5S 20.4–21 h 95.7 72 1426-05 2455272.7 3–11S 22–23 h 98 97 7.1 1428-04 2455274.7 3–11S 22–23 h 97.2 84 16.3 1774-12 2455620.7 1S–10N 0.5–1 h 93.5 96 9.5 1774-14 2455620.7 3–15N 0.2–1 h 92.7 110 16.6 1914-03 () 2455760.6 28–33N 0–1.3 h 94.9 83 13.3
Data both in the visible and IR ranges
317-06 2454163.8 11–30N 0.1–0.7 h 94.7 111 8.8 317-07 () 2454163.8 23–44N 0–0.4 h 95.5 65 8.9 323-07 () 2454169.8 25–50N 0.4–1 h 96.2 53 8.2 324-06 () 2454170.8 11–30N 0.7–1.5 h 95.1 91 16.5 324-07 () 2454170.8 23–50N 0.5–1.2 h 95.8 57 9.7 327-06 () 2454173.8 25–50N 1–1.5 h 94.6 27 715-02 2454561.7 5S–9N 22.5–23.3 h 98 92 11.2
Fig. 1. Average spectrum of VIRTIS in the 350–800 nm spectral range. The identified Herzberg II and Chamberlain transitions are labelled by the (0–m00) labels above and below
the spectrum line, respectively.
Fig. 2. Examples of tangent profiles of the Herzberg II system. They are identified by the corresponding VIRTIS orbit number and session, as reported inTable 1. They are obtained by considering the contribution of the single bands withm00= 7–11 of the
system. The continuum is estimated as described in Soret et al. (2010). The uncertainty in the position is assumed to be equal to 3 km, that is the vertical bin size.
peak altitude, the intensity at peak, and the full width at half max-imum (FWHM). The tangent altitude is in the range 92.7–98 km, with a FWHM ranging from 7.1 to 16.5 km. However, for some images, the SNR is very low and the fit with the Gaussian curve is difficult. For these cases, namely the 327-06, 1288-05, 1290-04, 1377-03, the FWHM is not derived. Considering the other cases, the FWHM is equal to 11.5 ± 3.2 km on average.
The intensities, reported inTable 1, are quite variable. In some cases, marked with a star, the Herzberg II progression is only partly detectable. For these cases, the reported intensity is not represen-tative because the contribution from some bands is missing, and hence not included in the discussion. In the case of image 1290-04, marked with ( ) inTable 1, the total intensity is overestimated, because of problems in the definition of the (0–9) transition band. This image is not included in the discussion. In the present work, we consider as representative intensities only those data where the (0–7) up to the (0–11) bands were simultaneously detected. For these cases, the total integrated intensity along the line of sight ranges between 84 kR and 116 kR. Considering the intensity factor at limb equal to 50, as mentioned in Section2, the observed Herz-berg II intensities are in good agreement with the values reported in the Venera data (Krasnopolsky, 1983). The uncertainty on the absolute intensity of the single measurements is estimated to be on the order of 20%. The peak altitude of the emission is not vari-able from one band to the other, as mentioned above. Hence, although in several cases the complete progression is not observed, we can derive a statistical behaviour of the peak altitude of the Herzberg II observed bands, as shown inFig. 3.
The mean value of the peak position is 95.5 ± 1.6 km, which coincides also with the weighted average. Despite the low number of profiles where the emissions in the visible spectral range can be studied, the distribution shows several similarities with the char-acteristics of the (0–0) O2 IR atmospheric band (Piccioni et al.,
2009). In that case, the statistical distribution of the peak position in the case of the (0–0) O2band was discussed, as retrieved with
the onion peeling technique. The mean value in the case of the tan-gent peak altitude was also reported, equal to 95 ± 2.4 km. The comparison between the Herzberg II and the (0–0) O2band is
dis-cussed in detail in Section3.2, where simultaneous observations in the visible and IR are presented.
3.1. Chamberlain system
Three bands of the Chamberlain system were also detected. They correspond to the 0–
m
00progression withm
00= 6, 7, 8, centeredat 558 nm, 604 nm, and 657 nm. However, because of the low SNR at wavelengths longer than 640 nm, we were able to investigate the limb profile only for the (0–6) and (0–7) bands. In the images named 1426-05 and 1774-12 inTable 1, the two bands are well
defined. In the case of image 1426-05, the intensities are equal to 9.8 kR and 10.3 kR respectively for the (0–6) and (0–7) bands, while in the image 1774-12 the intensities are 6.8 kR and 8.6 kR for the same bands (seeFig. 4). The intensity ratio I(0–7)/I(0–6) is estimated to be 1.05 and 1.27 for the two images respectively; these values are in a good agreement with the theoretical intensity ratio of 1.28 (Bates, 1989). However, a better statistics is required to derive firm conclusions. The total integrated intensity profile, obtained considering the (0–6) and the (0–7) bands of the Cham-berlain system from image 1774-12, is shown with a dashed line inFig. 4as an example.
The profile of the Chamberlain bands is also compared to the to-tal intensity of the Herzberg II system measured simultaneously. It appears that the Chamberlain system peaks a few km higher than the Herzberg II system, though this is less evident in the case of im-age 1426-05. This behaviour is discussed in relation with a one-dimensional model applied to these observations (Gérard et al., 2012). Unfortunately, the statistics of the Chamberlain bands is very poor, as a consequence of the weakness of these bands and also to their vicinity to the Herzberg II bands. Hence, it is quite dif-ficult to separate them from the background in most of the avail-able data.
According to the Venera observations, the Chamberlain band emissions are on average 13 times less intense than the Herzberg II bands (Krasnopolsky, 1983). In the observed cases, the Chamber-lain (0–6) and (0–7) bands are respectively 10 times and 9.4 times less intense than the integrated Herzberg II system intensity, con-sidering a peak intensity of 97 kR for the image 1426-05, while they are 14.1 times and 11.1 times less intense than the Herzberg II system intensity, assuming a peak intensity of 96 kR for the Herzberg II system, for the image 1774-12. However, we must take into account that these values only refer to two bands of the Cham-berlain system.
Concerning the peak altitude, according toKrasnopolsky’s mod-el (2011), the Chamberlain system has a maximum of emission at 110 km. In the VIRTIS data, we find that the Chamberlain bands emissions occur at about 100 km tangent height. Considering the difference between the tangent height and the altitude of the Vol-ume Emission Rate, the former being 2–5 km lower than the latter, we can conclude that there is quite good agreement between the expected altitude from this model and our results, in particular that the Chamberlain system lies above the Herzberg II one. This is further confirmed by an ad hoc one-dimensional atmospheric model developed to interpret our results, based on neutral densi-ties derived from Venus Express remote sensing observations. A brief discussion of the model and its conclusions is reported in Sec-tion4, while a more detailed description of the model can be found
Fig. 3. Statistical distribution of the Herzberg II peak altitude, acquired in the period from 2007-03-04 to 2011-07-18. The mean value of the peak position is 95.5 ± 1.6 km, which corresponds also to the weighted average.
Fig. 4. Tangent profile of the integrated intensity obtained by considering the (0–6) and the (0–7) bands of the Chamberlain system, compared to the tangent profile of the Herzberg II band system. The example refers to the VIRTIS image 1774-12 acquired on 2011-02-28.
inGérard et al. (2012). Noteworthy the results reported here are the first observations of the Chamberlain emission altitude profiles at Venus.
3.2. Comparison with O2infrared emissions
Details of the data included in this investigation are listed in Ta-ble 1. Data were acquired in the period from 2007-03-04-03 to 2011-07-18. For the data acquired from 2007-03-04 to 2008-04-05, reported in the lower part ofTable 1, both visible and IR spectra are available. The cases shown are the best ones where the 0–
m
00progression of the Herzberg II bands are clearly observed. This might be because the O2airglow emissions were stronger than in
the other cases, or because the SNR was good. For these data, a di-rect comparison of the O2nightglow emissions in the visible and IR
ranges in the Venus atmosphere is possible. Only two cases of di-rect comparison between visible and IR nightglow emissions are shown inFig. 5as an example. Data were acquired on 2007-03-04 (Fig. 5a) and 2008-04-05 (Fig. 5b). However, a few other cases, where simultaneous O2nightglow emissions in the visible and IR
are detected, were analyzed and listed inTable 1.
The intensity of the O2emission in the IR was scaled by a factor
of 0.002, in order to have the two curves in the same plot. A good agreement between the peak altitude of the Herzberg II and the (0– 0) emission peak of the O2ða1Dg—X3RgÞ is found, as seen inFigs. 5a
and b. For the particular cases analyzed in this work, the Herzberg II system intensity is about 500 times less intense than the O2ða1Dg—X3RgÞ (0–0) band in the infrared.
For the seven images listed in the lower part ofTable 1, where the IR and visible spectra were acquired simultaneously, we com-pare the peak altitude at limb for the two spectral ranges (Fig. 6). The same trend of the peak altitude fluctuation is observed in the
IR and visible. The peak positions are compatible within the statis-tical errors, for the IR (0–0) band and the Herzberg II bands, though the latter peak is on average slightly lower than the (0–0) band. On average, the (0–0) emission peak of the O2ða1Dg—X3RgÞ is observed
at 97.4 ± 2.5 km, as reported in the statistical study of this emission inPiccioni et al. (2009), based on VIRTIS/Venus Express data, and the relative peak position at limb is on average at 95 ± 2.4 km. The FWHM is variable with latitude, ranging from 11 km at equator to 6 km at mid-high latitudes (Piccioni et al., 2009). In the case of the Herzberg II bands, the peak emission is observed at about 95–96 km in limb view, while the FWHM is wider than the mean value for the IR emission bands. Considering the high variability of this parameter and the limited sampling in latitude of the data in the visible range, which are concentrated close to equator where the FWHM is equal to 11 km in the IR, a similar value for the FWHM is found. The average trends are in good agreement. How-ever, the behavior with respect to latitude in the case of the Herz-berg II bands cannot be investigated at this time because of an insufficient coverage at high latitudes.
4. One-dimensional simulations
The results reported here allow refining the modelling of the Venus nightside airglow. Based on the recent VIRTIS observations of the O2 nightglow emissions, both in the visible and infrared
spectral regions, it is possible to constrain the production and col-lisional deactivation of excited O2molecules. The model is based
on a steady state equation for the population of the upper state of the transition (Gérard et al., 2012). The atomic oxygen and CO2 densities used in the model are taken from previous VIRTIS
and SPICAV measurements (Soret et al., 2012). According to this calculation, the peak position of the Herzberg II bands is located slightly lower than the peak position for the a1D(0–0) band in
the IR, as it is observed in the simultaneous VIRTIS data in the vis-ible and IR (seeFig. 6). The Chamberlain bands are predicted to peak about 4 km higher than the Herzberg II bands. This confirms the few VIRTIS observations where the altitude of the Chamberlain bands may be measured, a consequence of the faintness of these emissions (see Fig. 4). The 4–5 km altitude difference between the Infrared Atmospheric and the Herzberg II bands on one hand, and the Chamberlain bands on the other hand, is presumed to be a consequence of smaller quenching rate coefficients for the a and c singlet states compared to more efficient deactivation of the A0triplet state by CO
2molecules and O atoms.
Fig. 5. (a) Comparison between the vertical profile of the O2emissions in the IR at
1.27lm and in the visible range. The intensity of the emission in the IR was scaled to a factor 0.002, in order to make easier the comparison with the Herzberg II bands intensity and have the two profiles in the same plot. The IR and visible spectra are simultaneously acquired on 2007-03-04 (image number 317-06). (b) The same for the VIRTIS images acquired on 2008-04-05 (image number 715-02).
Fig. 6. Peak altitude at limb for the Herzberg II integrated band system (marked with a triangle), as derived in this work, compared to the peak altitude of the (0–0) emission of the O2ða1Dg—X3RgÞ (marked with a round symbol), fromPiccioni et al.
(2009).
A comparison between observed and predicted relative ties shows that the model is able to reproduce the relative intensi-ties of the Herzberg II and the (0–0) IR Atmospheric transition satisfactorily, while the agreement between model and observa-tions is worse for the Chamberlain bands, possibly due to the lim-ited number of observations.
5. Conclusions
We report the detection of eight bands of the O2Herzberg II and
three bands of the Chamberlain systems in the visible spectral range, observed by the VIRTIS spectrometer on board Venus Express.
The good sampling in the altitude range of about 90–120 km, achieved with the limb-tracking technique, allows investigation of the vertical profile of the detected bands. We found that the two systems peak at different altitudes. In particular, the Chamberlain bands peak about 4 km higher than the Herzberg II bands. This trend is well reproduced in the emission model devel-oped using oxygen and CO2profiles retrieved from Venus-Express
observations.
The integrated intensity of the Herzberg II system, with
m
00= 7–11, varies in the range 84–116 kR, in agreement with the previous detection byGarcía-Muñoz et al. (2009). For the Chamberlain sys-tem, only two bands, namely (0–6) at 558 nm, and (0–7) at 604 nm were investigated in this work. For these bands, the relative inten-sity ratio I(0–7)/I(0–6) has been calculated and compared with the theoretical value, equal to 1.28. The results are in reasonable agree-ment with theory. The (0–6) Chamberlain band falls very close to the so-called atomic oxygen green line, which occurs at 557.7 nm. However, no indication of the latter has been found in the VIRTIS data so far. This could be a consequence of the faintness or the sporadic presence of the oxygen green line, or this could be explained by the fact that the altitude we are considering for the Chamberlain band detection is not where the green line emission occurs, which is still debated. A further investigation of the VIRTIS data at limb might shed more light on the green line detection.
The results presented allow us to improve our knowledge about the minor species in the Venus atmosphere. With the information derived from the vertical profiles of these band systems, we have a great improvement of the knowledge of the vertical structure and hence of the dynamical properties of the night side atmosphere of Venus. Moreover, the Chamberlain bands, peaking at an altitude different from Herzberg II and the IR atmospheric bands, allows tracing of the day-to-night circulation at different altitudes, lead-ing to a 3-D structure of the atmosphere, as has already been done, by combining the ultraviolet NO and IR O2emissions (Gérard et al.,
2009).
Acknowledgments
The authors thank ESA, ASI, CNES and all the national space agencies which support Venus-Express. AM is supported by ASI (Grant: ASI-INAF I/050/10/0). TGS supported by NASA Planetary Atmospheres Grant NNX08AO27G. Partial funding for this research
was provided by the PRODEX program of the European Space Agency, managed in collaboration with the Belgian Federal Science Policy Office.
References
Baker, D.J., 1974. Rayleigh, the unit for light radiance. Appl. Opt. 13, 2160–2163. Bates, D.R., 1989. Oxygen band system transition arrays. Planet. Space Sci. 37, 881–
887.
Bougher, S.W., Borucki, W.J., 1994. Venus O2visible and IR nightglow: Implications
for lower thermosphere dynamics and chemistry. J. Geophys. Res. 99, 3759– 3776.http://dx.doi.org/10.1029/93JE03431.
Cardesín Moinelo, A., Piccioni, G., Ammannito, E., Filacchione, G., Drossart, P., 2010. Calibration of hyperspectral imaging data: VIRTIS-M onboard Venus Express. IEEE Trans. Geosci. Remote Sens. 48, 3941–3950.
Cosby, P.C., Sharpee, B.D., Slanger, T.G., Huestis, D.L., Hanuschik, R.W., 2006. High-resolution terrestrial nightglow emission line atlas from UVES/VLT: Positions, intensities, and identifications for 2808 lines at 314–1043 nm. J. Geophys. Res. 111, A12307.
Drossart, P. et al., 2007. Scientific goals for the observation of Venus by VIRTIS on ESA/Venus Express mission. Planet. Space Sci. 55, 1653–1672.
García-Muñoz, A., Mills, F.P., Slanger, T.G., Piccioni, G., Drossart, P., 2009. Visible and near-infrared nightglow of molecular oxygen in the atmosphere of Venus. J. Geophys. Res. 114, E12002.http://dx.doi.org/10.1029/2009JE003447.
Gérard, J.C. et al., 2009. Concurrent observations of the ultraviolet nitric oxide and infrared O2nightglow emissions with Venus Express. J. Geophys. Res. 114,
E00B44.
Gérard, J.C., Soret, L., Migliorini, A., Piccioni, G., 2012. Oxygen nightglow emissions of Venus: Vertical distribution and role of collisional quenching. Icarus.http:// dx.doi.org/10.1016/j.icarus.2012.11.019.
Greer, R.G.H., Murtagh, D.P., 1985. Singlet oxygen nightglow in the atmosphere of Earth, Venus and Mars. In: Proceedings of Galway A.S.G.I. Meeting.
Krasnopolsky, V.A., 1983. Venus spectroscopy in the 3000–8000 Å region by Veneras 9 and 10. In: Hunten, D.M. et al. (Eds.), Venus. Univ. of Ariz. Press, Tucson, pp. 459–483.
Krasnopolsky, V.A., 2011. Excitation of the oxygen nightglow on the terrestrial planets. Planet. Space Sci. 59, 754–766.
Krasnopolsky, V.A., Krysko, A.A., Rogachev, V.N., Parshev, V.A., 1977. Spectroscopy of the night-sky luminescence of Venus from the interplanetary spacecraft Venera 9 and 10. Cosmic Res. 14, 687–692.
Lawrence, G.M., Barth, C.A., Argabright, V., 1977. Excitation of the Venus night airglow. Science 195, 573–574.
Piccioni, G. et al., 2009. Near-IR oxygen nightglow observed by VIRTIS in the Venus upper atmosphere. J. Geophys. Res. 114, E00B38. http://dx.doi.org/10.1029/ 2008JE003133.
Piccioni, G. et al., 2007. VIRTIS: The Visible and Thermal Imaging Spectrometer. ESA-SP 1295, ESA Publications Division, Noordwijk, The Nederlands.
Slanger, T.G., 1978. Generation of O2ðc1Ru;C3Du;A3RþuÞ from oxygen atom
recombination. J. Chem. Phys 69, 4779–4791. Slanger, T.G., Black, G., 1978. The O(1
S) airglow – new laboratory results. In: Kurylo, M.J., Braun, W. (Eds.), 12th Informal conference of Photochemistry, Gaithersburg, MD, USA, 28 June–1 July 1976. Natl. Bur. Standards, Washington, DC, USA, pp. 102–104.
Slanger, T.G., Copeland, R.A., 2003. Energetic oxygen in the upper atmosphere and the laboratory. Chem. Rev. 103, 4731–4765.
Slanger, T.G., Cosby, P.C., Huestis, D.L., Bida, T.A., 2001. Discovery of the atomic oxygen green line in the Venus night airglow. Science 291, 463–465. Slanger, T.G., Cosby, P.C., Huestis, D.L., 2003. A new O2 band system: The
c1R u—b
1 Rþ
g transition in the terrestrial nightglow. J. Geophys. Res. 108 (A2),
1089.http://dx.doi.org/10.1029/2002JA009677.
Slanger, T.G., Huestis, D.L., Cosby, P.C., Chanover, N.J., Bida, T.A., 2006. The Venus nightglow: Ground-based observations and chemical mechanism of the oxygen airglow. Icarus 182, 1–9.
Soret, L., Gérard, J.C., Piccioni, G., Drossart, P., 2010. Venus OH nightglow distribution based on VIRTIS limb observations from Venus Express. Geophys. Res. Lett. 37, L06805.http://dx.doi.org/10.1029/2010GRL042377.
Soret, L., Gérard, J.C., Montmessin, F., Piccioni, G., Drossart, P., Bertaux, J.L., 2012. Atomic oxygen on the Venus nightside: Global distribution deduced from airglow mapping. Icarus 217 (2), 849–855.
